The officers were elected at this annual meeting, and the individual members were proposed as Fellows, Honorary, Corresponding, and Foreign Members.

By an act approved Feb. 5, 1841, the American Statistical Association was incorporated by the legislature of Massachusetts. Thus the new Association, full-fledged and authorized by law, was ready for its work of preserving and diffusing statistical information. It has always adhered to this provision of the act incorporating it. It has not gone into economic or social questions, philosophical or ethical science, but has adhered most rigidly to its statistical objects.

Up to date it has had a continued existence, there never having been a year since its organization that it has not held meetings and had papers of a statistical character, although its proceedings have not been regularly printed. It has had during the whole period of its existence of sixty-nine years but five Presidents: Hon. Richard Fletcher, 1839-45; George C. Shattuck, M.D., 1846-51; Dr. Edward Jarvis, 1852-82; Dr. Francis A. Walker, 1883-96; and the present incumbent, 1897 to date.

Dr. Jarvis served the Association thirty years as its President, and on the election of his successor, Dr. Walker, he was made President Emeritus.

With two exceptions three quarterly meetings, besides the annual meeting, each year were held from 1840 to 1879. Some of the meetings were omitted in some of the years, especially the July meeting, which, beginning in 1881, was omitted regularly. From 1894 on no quarterly meetings were held, with the exception of a special quarterly meeting on April 16, 1897, in memory of President Walker. Since 1899 until the present year no quarterly meetings have been held, but there has been no year in which meetings have been entirely omitted.

The proceedings of the society have not been published regularly. In 1847 the collections of the Association were brought together and printed, entitled Volume I., in three parts. In the first part there appears quite a valuable paper by Professor B. B. Edwards, of the Andover Theological Seminary.

Professor Edwards discussed the history and origin of statistics. There were other papers on the towns of Massachusetts in this first part, with a history of their origin.

Part two contained statistics of population in Massachusetts, prepared by Rev. Joseph B. Felt, while part three, published in 1847, contains statistics of taxation in Massachusetts, including valuation and population. This was also prepared by Mr. Felt, and was one of the first attempts at analysis of statistics by a member of the Association.

So far as I have been able to learn, papers in a desultory way were published from time to time, but no regular collection appeared until what we now know as the New Series, beginning in March, 1888, under the direction of Dr. Davis R. Dewey, and when General Walker was President of the Association. Since then the publications have appeared regularly, and they constitute a collection of exceedingly valuable statistical productions. In fact, I feel warranted in asserting that no statistical society has, on the whole, brought out a more valuable collection of statistical material, and the members of the Association can feel gratified that it has presented to the public so many carefully analyzed topics relating to the science to which we are devoted.

With this brief historical review, it is pertinent to discuss the field of labor of the Association, the condition of public statistics when it was organized, and the opportunity it has had for exerting its influence and for conducting its statistical investigations.

The scope of statistical inquiry when the Association was organized was not only very limited in quantity, but meagre and unsatisfactory in quality. The vehicle of statistical information, the federal census, had not reached encyclopedic proportions. The Association had as its field from which to draw the facts for its analysis the United States census from 1790 to 1840, the Colonial censuses, and those of one or two States, more especially the efforts of the Commonwealth of Massachusetts. A glance at the founders of our Association convinces one immediately that they were men who understood and com

prehended both analysis and classification; that the analysis and classification of the facts relating to various conditions which surround the human race entered into their great object, and the purpose of social science at large, the chief object of which is to spread the knowledge resulting from the investigations of its movements, that the people may better appreciate and understand their own conditions and aid, by an increased intelligence, in the amelioration of unfavorable features and the eradication of positive evils.

They knew as well as social scientists and statisticians to-day that statistical methods-or statistical science, if you prefercould evolve laws which should be applied practically to these great objects of social science. They felt that the time had arrived when something should be done outside the mere collection of data. They had an opportunity with the meagre efforts back of them for this purpose.

The United States instituted the national census in 1790. There are three periods to the American census, the Colonial, the Continental, and the Constitutional, or our present period. During the first period the British Board of Trade played an important part in American affairs, and it often attempted enumerations of the people of the colonies, but the census had not then assumed scientific form and definiteness in Europe, and, as would be expected, the results here were very imperfect. Superstition was an obstacle, but without obstacle success could not have been attained in the colonies when the mother country took her first census in 1801, and then so imperfectly. that the results were of no immediate value.

During the Continental period, although resolutions in Congress had been introduced, no general enumeration of the population was secured. Various estimates and computations were made from time to time, but they came no nearer accuracy than those made in the Colonial period. It had, however, become clearly settled that there never could be a complete enumeration until the work was done by a central directing authority. It was left to the Constitution to give us first an enumeration

of population and afterwards a national census, primarily for the purpose of apportioning representatives and direct taxes among the several States included within the Union and according to their respective numbers.

This attitude of Congress caused an enumeration of the population in 1790, and from this has grown the national census. Perhaps the most impressive statement relative to this growth relates to the number of inquiries at the first and at later cen

suses.

At the census of 1790 there was one schedule, containing four inquiries. In 1840, when the American Statistical Association had just been organized, there were two schedules, containing 82 inquiries or details. In 1890, just one hundred years after the first census, there were 233 schedules, containing 13,161 inquiries or details. Reduced in 1900 to 7,476. Of course, in 1790 the 4 queries related to the members of the family, the people only, while in 1890 the inquiries did not apply to one individual, but they were all projected. This was the grand sweep of one hundred years.

In 1810 an attempt was made to collect data relative to manufactures. This was repeated in 1820, omitted in 1830, and taken up again in 1840, and has continued through all censuses since that time, but until 1850 the inquiries as to manufactures amounted to but little.

A start was made in agricultural statistics in 1840, and the work has been continued throughout.

There were Colonial censuses in the colony of Massachusetts in 1754-1765-1776. State censuses were ordered in 1837-1840 and 1850. The regular decennial enumeration of the inhabitants under State authorization was ordered taken in 1855 in connection with the collection of industrial statistics, and this has been taken since then, being the mean between the dates of the federal census. Thus Massachusetts has a census every five years, both of population and of industry. Some other States organized a census on the quinquennial period relative to the United States census, but, like the earlier State censuses of Massachusetts, they were of very little value except in se

curing the aggregate population for the purpose of legislative representation.

The crudeness of these earlier censuses seems very strange to us, and yet they were creditable efforts on the part of the State governments authorizing the collection and showed a disposition to secure information on which to base conclusions and actions.

The first analytical survey of any of these works, so far as my observation warrants the statement, was by Lemuel Shattuck, the first Secretary of this Association, in a report to a committee of the city council appointed to obtain the census of Boston for the year 1845.

Mr. Shattuck, who drew this report for the committee, indulged in some very sharp criticisms of the federal census, and analyzed in a very creditable way the results of that particular census. Some of the members of the Association-and I regret that I have not been able to find the documentsdealt with the State censuses from time to time, and others, especially Dr. Jarvis, had much to do with inducing the federal government to expand its work, and in 1870 he analyzed some of the statistics of the federal census.

I do not know whether any of you ever knew Dr. Jarvis, but I knew him well, not only through my association with him in this organization, but by his frequent visits to my office to discuss statistical questions. He was very fond of relating some of the amusing things he found in the federal census. On one occasion he found the case of a man something over eighty years of age who had died of teething, and a child a few months old who had died of old age. Of course there was a transposition in the statements, but it amused the old doctor, immensely and he never tired of relating the anecdote.

The first census, as I have said, which really amounted to an attempt at scientific work was the federal census of 1850. It was better in some respects than the succeeding one of 1860. In 1870 General Walker was put in charge, and the census of that year was an enormous improvement over that of 1850; but it was just prior to 1880 that General Walker submitted